"Water"-cooling Question

Firstly, I know close on zilch when it comes to watercooling. That said, I've been wondering for quite some time now - what would happen if one were to use pure alcohol instead of water in their system? Call it idle curiosity. :)

er, it would flow around your system being much less efficient at conducting heat away from your waterblocks than water...

What makes you think alcohol would be anything special?

Flourinert, or similar would be more interesting, since the PC can be immersed, but I haven't seen a long term setup using this method.

Like tp says the heat conductivity is not as good as water.

There are other liquids that can be used but non of them seem to be as good (heat wise) as water.

The main advantages of things like fluidxp is that they don't need changing frequently.

Well, apart from having a far lower freezing point than water, it also evaporates at a comparatively low temperature. Assuming that it would, evaporating and re-condensing would use up heat energy. Then again, I'm under the impression that having bubbles in your system is a bad thing. *shrugs* Just one of my more strange ideas. :)

The only reason i would use a liquid other than water is if i wanted to cool it to a very low temperature.

Use a liquid that evaporates at 40, that would work resonably well as you can take advantage of the heat it will absorb as it changes states (temp doesnt rise as this happens)

The best reson to use alcohol is if its non conductive (although many are).

Specific heat is not all that matters in a coolant. Water has the best specific heat that I know of. The viscosity of many liquids is higher than water and so makes it much harder to pump. Also some will degrade tubing and the pump, possibly also affecting the metal in the system aswell.
The surface tension does also affect performance a lot - becuase water clings to the sides of its container, there is a layer of coolant that either does not move, or moves very slowley in comparason to the rest of the water. Most recentblock designs have been aimed at reducing this boundry layer and getting as many particles of water as possible to touch the hot parts of the block, in an effort to increase the effectiveness of the surface area.

The thing is... why to do that???
Years now all of us use water with some kind of additive and it works perfect... using a different liquid will achieve what, ~1-2 degrees less??? Buy a better waterblock and use good thermalpaste if you want better temps and stability ;)

My first point would be a very goodidea, as it borders on extreemly cheap phase change, although you would need to take some precoutions to stop gas from getting in the pump....

There have been many debates about coolant fluid and so far they have all gave the conculsion that noone can be arsed to do the research. I certinally cant. Water is good enough for me, add about 5ml of industrial strength biocide and some purple ice (which doesnt reallly do anything but its in there anyway!) and you shouldnt have any problems except for leaks which rearly kill anything but can be quite annoying.

True... it is pretty pointless to do such a research at least for watercooling pc components anyway.

So, were one to go for this, the radiator and resevoir would go somewhere up top, and the pump at the bottom of the case, below the waterblock(s)? Or is this a general rule of thumb? In fact, are there any good places you'd recommend I read up on watercooling in general? My curiosity's been piqued. :) Which probably means I'm not gonna be adding to my savings, then. :(

forums.procooling.com
Has everything about watercooling hidden somewhere behind the search function :)

positioning doesnt really matter for normal w/c, although for obvious resons you may want the res at the top or near the top to help with filling...

my rads at the bottom, sucking through the bottom of the case, that way I dont have nasty air holes showing anywhere and I can hide a dust filter there too...

As your in SA you might want to look at DIY w/c, if you can get all the stuff there instead of shipping it half way round the world it could be a lot cheaper :)

Methanol is expensive, however it has boiling point of 65c, so that might be resonable.
The problem is that all low-temp alcohols also have very low flash points so they would either have to be mixed with plenty of water (in effect raising the boiling point considerably making the whole idea pointless) or the system must be purged from having any oxygen atall and be leak free. That wouldnt acctualloy be too hard to do though. Fill the system with the alcohol, then using a sherader (sp?) valve spray in a denser gas than air with the filling hole slightly open, butane or possibly nitrogen (I have plenty of liquid nitrogen) would do nicly (and I have plenty if I were to try it) and then seal the filling valve, there would be no air left :)

We're not entirely isolated here in South Africa - one can get Thermaltake and Cooler Master water cooling stuff here. With more work, Prommies and such are also to be found. Still, importing wouldn't be too much of an issue - my CPU (Athlon Mobile, non-existent here), Raddy (Too expensive here) and Samsung harddrives (Discountinued around the time 3GB was considered huge) are all imported, albeit my previous importing source (My aunt ;)) is now discountinued.

Anyway. More and more, I'm wishing I paid more attention to organic chemistry. Though redox reactions were rather boring. :P Going with a compound with a lower boiling point would be a bad idea, given that you'd have to ensure the system is pressurized, right? Also, would methanol actually reach its boiling point in an average system (On air, I've never let my CPU over 60 C)?

thermal take and coolermaster are worse than air cooling though :P

nope, wouldnt have to be pressureised, just leak free, although if it did evaporate, that would create a small pressure on the system.
You would only need pessure if you were using a very low temp refrigerant, so that it can be compressed into a liquid.
As I said though, the problem is there flammable (and more expensive), which could potentially be quite a large problem.

My point was, we're not entirely isolated. :P

Seem to have been thrown in the deep end with the forums you mentioned. That, or I just seem to stumble into all the technical threads (RTDs? Thermocouples? *hides in fear*).

Thanks for your time. :)

heh :P

also forums.bit-tech.net has some w/c stuff aswell